binomial model fixed effect parameter estimates reliable?
Posted: Mon May 10, 2010 11:42 pm
In the MLwiN FAQs under discrete response models it states that
..."It is known that quasi-likelihood methods give estimates for the random parameters which are biased downwards (see RodrÃguez and Goldman, 2001). For this reason, we do not recommend that users take results from these estimation methods as final, but instead recommend that they should use MCMC estimation (after using quasi-likelihood estimation to get starting values). "
Does anyone know if the standard deviation estimates for the fixed effect parameter estimates are ok to use for testing significance of, say a treatment effect?
Thanks very much!
..."It is known that quasi-likelihood methods give estimates for the random parameters which are biased downwards (see RodrÃguez and Goldman, 2001). For this reason, we do not recommend that users take results from these estimation methods as final, but instead recommend that they should use MCMC estimation (after using quasi-likelihood estimation to get starting values). "
Does anyone know if the standard deviation estimates for the fixed effect parameter estimates are ok to use for testing significance of, say a treatment effect?
Thanks very much!